Hybrid Digital Image Watermarking using Contourlet Transform (CT), DCT and SVD
Abstract
Role of watermarking is dramatically enhanced due to the emerging technologies like IoT, Data analysis, and automation in many sectors of identity. Due to these many devices are connected through internet and networking and large amounts of data is generated and transmitted. Here security of the data is very much needed. The algorithm used for the watermarking is to be robust against various processes (attacks) such as filtering, compression and cropping etc. To increase the robustness, in the paper a hybrid algorithm is proposed by combining three transforms such as Contourlet, Discrete Cosine Transform (DCT) and Singular Value Decomposition (SVD). Performance of Algorithm is evaluated by using similarity metrics such as NCC, MSE and PSNR
Keywords: Digital Image Processing, Discrete Wavelet Transform, Discrete Cosine Transform, Contourlet Transform, Hybrid Water Marking..
1 Introduction
Digital watermarking is an emerging area in computer science, digital signal processing and communications. It is intended by its developers as the solution to the problems of data copyright, content protection and ownership proof. Digital watermarking is the process of embedding a piece of information into the digital data (audio, image or video), which can be detected and extracted later to make an assertion about the data. Recently, several transforms are proposed to represent the image and watermark in a small number of coefficients.
Hybrid watermarking is a method where more than one transform is used to generate image and watermark coefficients are more potential than a single transform method. Watermarking with single transform is robust or semi-fragile with limited and bounded. Hybrid watermarking with multiple transforms is a more robust and semi-fragile.
2. Background
Watermarking can be implemented either in the time or frequency (transform) domain. Transform domain watermarking techniques apply some invertible transforms to the host image before embedding the watermark. Then the coefficients are modified to embed the watermark and finally the inverse transform is applied to obtain the watermarked image. The watermark embedded in the transform domain is irregularly distributed over the area and make more difficult for an attacker to extract or modify the watermark. The transforms commonly used for watermarking are hybrid Image Watermarking Methods are combining two or three transforms like DWT - DCT, DCT - SVD, DWT - SVD, DWT- DCT-SVD, Contourlet – DCT, and Contourlet – SVD ..etc. In this paper hybrid watermarking method combines three transforms such as Contourlet , DCT, and SVD are imperceptible and robust.
2.1 Contourlet Transform (CT)
The Contourlet Transform is a geometrical transform, can efficiently detect image edges in all directions. It is widely used in various signal processing applications, including image watermarking. CT consists of two major parts, the Laplacian Pyramid (LP) and Directional Filter Bank (DFB) as shown in FIGURE1. The Laplacian Pyramid (LP) is constructed from a pair of filters known as Analysis and Synthesis filters as shown in FIGURE2(a) and (b) respectively.
2.2 Discrete Cosine Transform (DCT)
A DCT represents the input data points in the form of a sum of cosine functions that are oscillating at different frequencies and magnitudes. There are mainly two types of DCT: one dimensional (1-D) DCT and two dimensional (2-D) DCT. Since an image is represented as a two dimensional matrix, for this research work, high compaction 2-D DCT is considered.
2.3 Singular Value Decomposition (SVD):
Every real matrix A can be decomposed into a product of 3 matrices A = UDVT, where U and V are orthogonal matrices. The diagonal entries of D are called the singular values of A, the columns of U are called the left singular vectors of A, and the columns of V are called the right singular vectors of A. This decomposition is known as the Singular Value Decomposition (SVD) of A
3. Proposed method:
Digital watermark is an embedded (hidden) marker in digital multimedia data by watermarking method, the marker is generally unobservable, which can be extracted by special detector. The basic idea for digital watermark is to use human‘s insensitive perceptual organs and redundancy in digital signal and embed secret information in digital products, such as image, audio frequency and video frequency in order to easily protect its copyright, and in addition, the embedded information survives after attacks so as to perform copyright authentication and protection. Digital watermark doesn‘t change the basic characteristic and value of the products. Watermarking system is consists of two parts, 1) watermark creation and embedding 2) watermark extraction. The following steps are used to implement the watermarking:
Watermark embedding process
Watermark Extraction process
Performance Evaluation
3.1. Watermark Embedding Process
Before embedding the watermark in to the original image it will be transformed into coefficients by applying CT, DCT, and SVD. Original image also transformed into coefficients and then both are applied to the embedding algorithm which is known as watermarked image, now inverse transforms are applied to obtain the spatial domain watermarked image as shown in FIGURE 3.
Figure 3: Watermark Embedding process.
3.3. Watermark Extraction Process
Watermark extraction process deals with the extraction of the watermark in the absence of the original image as shown in FIGURE 4. The aim of the watermark extraction algorithm is to obtain the reliable an estimate of the original watermark from the watermarked image. The extraction process is inverse of the watermark embedding process.
Figure 4: Watermark Extraction Process.
3.5. Performance Evaluation
An important way of evaluating watermarking algorithms is to compare the amount of distortion introduced into a host image by watermarking algorithm. In order to measure the quality of the image at the output of the decoder, Mean Square Error (MSE) and Peak to Signal to Noise Ratio (PSNR) are used.
4. SOFTWARE AND HARDWARE REQUIREMENTS
Operating system : Windows XP/7.
Coding Language: MATLAB
Tool:MATLAB R 2012
SYSTEM REQUIREMENTS:
HARDWARE REQUIREMENTS:
System: Pentium IV 2.4 GHz.
Hard Disk : 40 GB.
Floppy Drive: 1.44 Mb.
Monitor: 15 VGA Colour.
Mouse: Logitech.
Ram: 512 Mb.
5. CONCLUSION:
From the simulation results it is observed that high quality image i.e. Watermarked image with high PSNR is obtained by embedding the watermark high level decomposition. With the increase in the density of variance of Gaussian noise the amount of noise induced in to the image is increased and these affected the quality of image and modify the watermark embedded coefficient of the image. With the increase in density/variance of the noise the PSNR values decreases and the robustness of the watermark is affected, in spite of huge noise addition the recovered watermark is still highly recognizable.
The watermarking algorithm sustains the cropping attack the watermark is highly recoverable even for the cropping block size of 256 x 256. We can say that even the image pass through different attacks such as Geometric, Adding noise, Filtering , the logo is get extracted perfectly. We can say that the embedding algorithm is robust. More over the logo is extracted if the location of embedded is known, so the embedding algorithm is secure. The image after watermarking is subjected to the low pass filtering and high pass filtering to remove the detailed coefficients. High pass filter removes the low frequency coefficients. Low pass filter is used to reduce the high frequency components.
The watermarking algorithm sustains the compression attack; the watermark is highly recoverable even for the compression block size of M x N. We can say that even the image pass through different attacks such as Geometric, Adding noise, Filtering, the logo get extracted perfectly. We perform further experiments to evaluate the robustness against more commonly used image processing attacks such as JPEG2000 compression, and DCT Compression.
References
[1] S. Azizi, M. Mohrekesh and S. Samavi."Hybrid image watermarking using local complexity variations."In Electrical Engineering (ICEE), 2013 21st Iranian Conference on, pp. 1-6.IEEE, 2013.
[2] B. Fadaeenia and N. Zarei. ―Hybrid DCT-CT Digital Image Adaptive Watermarking.‖ In DBKDA 2011: The Third International Conference on Advances in Database, Knowledge, and Data Applications.pp. 47-53. IARIA, 2011.
[3] A. Akter and M. A. Ullah. "Digital image watermarking based on DWT-DCT: Evaluate for a new embedding algorithm." In Informatics, Electronics & Vision (ICIEV), 2014 International Conference on, pp. 1-6. IEEE, 2014.
[4] S. A. Kasmani and A. Naghsh-Nilchi. "A new robust digital image watermarking technique based on joint DWT-DCT transformation." In Convergence and Hybrid Information Technology, 2008.ICCIT'08. Third International Conference on, vol. 2, pp. 539-544. IEEE, 2008.
[5] A. Al-Haj. "A hybrid digital image watermarking algorithm."In Innovations in Information Technology, 2007.IIT'07. 4th International Conference on, pp. 690-694. IEEE, 2007.
[6] F. Golshan and K. Mohammadi."A hybrid intelligent SVD-based digital image watermarking."In Systems Engineering (ICSEng), 2011 21st International Conference on, pp. 137-141.IEEE, 2011.